Buckets:

hf-doc-build
/

doc

Files

xet

hf-doc-build/doc / transformers /main /ko /tasks /asr.html

HuggingFaceDocBuilder

about 7 hours ago

download

raw

73.2 kB

	<meta charset="utf-8" /><meta name="hf:doc:metadata" content="{"title":"자동 음성 인식","local":"automatic-speech-recognition","sections":[{"title":"MInDS-14 데이터 세트 가져오기","local":"load-minds-14-dataset","sections":[],"depth":2},{"title":"전처리","local":"preprocess","sections":[],"depth":2},{"title":"평가하기","local":"evaluate","sections":[],"depth":2},{"title":"훈련하기","local":"train","sections":[],"depth":2},{"title":"추론하기","local":"inference","sections":[],"depth":2}],"depth":1}">
	<link href="/docs/transformers/main/ko/_app/immutable/assets/0.e3b0c442.css" rel="modulepreload">
	<link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/entry/start.acfca1c5.js">
	<link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/chunks/scheduler.53228c21.js">
	<link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/chunks/singletons.0cc5f498.js">
	<link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/chunks/index.e93d0901.js">
	<link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/chunks/paths.dab99b64.js">
	<link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/entry/app.f0c7a87a.js">
	<link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/chunks/preload-helper.cb103237.js">
	<link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/chunks/index.3db2ce32.js">
	<link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/nodes/0.3e3ee3d5.js">
	<link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/chunks/each.e59479a4.js">
	<link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/nodes/155.b12635b1.js">
	<link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/chunks/Tip.d62b0755.js">
	<link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/chunks/CopyLLMTxtMenu.1327b590.js">
	<link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/chunks/MermaidChart.svelte_svelte_type_style_lang.49b88d99.js">
	<link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/chunks/Youtube.b27f3a4a.js">
	<link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/chunks/CodeBlock.ada04ea6.js">
	<link rel="modulepreload" href="/docs/transformers/main/ko/_app/immutable/chunks/DocNotebookDropdown.1834da50.js"><!-- HEAD_svelte-u9bgzb_START --><meta name="hf:doc:metadata" content="{"title":"자동 음성 인식","local":"automatic-speech-recognition","sections":[{"title":"MInDS-14 데이터 세트 가져오기","local":"load-minds-14-dataset","sections":[],"depth":2},{"title":"전처리","local":"preprocess","sections":[],"depth":2},{"title":"평가하기","local":"evaluate","sections":[],"depth":2},{"title":"훈련하기","local":"train","sections":[],"depth":2},{"title":"추론하기","local":"inference","sections":[],"depth":2}],"depth":1}"><!-- HEAD_svelte-u9bgzb_END --> <p></p> <div class="items-center shrink-0 min-w-[100px] max-sm:min-w-[50px] justify-end ml-auto flex" style="float: right; margin-left: 10px; display: inline-flex; position: relative; z-index: 10;"><div class="inline-flex rounded-md max-sm:rounded-sm"><button class="inline-flex items-center gap-1 h-7 max-sm:h-7 px-2 max-sm:px-1.5 text-sm font-medium text-gray-800 border border-r-0 rounded-l-md max-sm:rounded-l-sm border-gray-200 bg-white hover:shadow-inner dark:border-gray-850 dark:bg-gray-950 dark:text-gray-200 dark:hover:bg-gray-800" aria-live="polite"><span class="inline-flex items-center justify-center rounded-md p-0.5 max-sm:p-0 hover:text-gray-800 dark:hover:text-gray-200"><svg class="sm:size-3.5 size-3" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg></span> <span>Copy page</span></button> <button class="inline-flex items-center justify-center w-6 max-sm:w-5 h-7 max-sm:h-7 disabled:pointer-events-none text-sm text-gray-500 hover:text-gray-700 dark:hover:text-white rounded-r-md max-sm:rounded-r-sm border border-l transition border-gray-200 bg-white hover:shadow-inner dark:border-gray-850 dark:bg-gray-950 dark:text-gray-200 dark:hover:bg-gray-800" aria-haspopup="menu" aria-expanded="false" aria-label="Open copy menu"><svg class="transition-transform text-gray-400 overflow-visible sm:size-3.5 size-3 rotate-0" width="1em" height="1em" viewBox="0 0 12 7" fill="none" xmlns="http://www.w3.org/2000/svg"><path d="M1 1L6 6L11 1" stroke="currentColor"></path></svg></button></div> </div> <div class="flex space-x-1 " style="float: right; margin-left: 10px; display: inline-flex; position: relative; z-index: 10;"> <div class="relative colab-dropdown "> <button class=" " type="button"> <img alt="Open In Colab" class="!m-0" src="https://colab.research.google.com/assets/colab-badge.svg"> </button> </div> <div class="relative colab-dropdown "> <button class=" " type="button"> <img alt="Open In Studio Lab" class="!m-0" src="https://studiolab.sagemaker.aws/studiolab.svg"> </button> </div></div> <h1 class="relative group"><a id="automatic-speech-recognition" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#automatic-speech-recognition"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>자동 음성 인식</span></h1> <iframe class="w-full xl:w-4/6 h-80" src="https://www.youtube-nocookie.com/embed/TksaY_FDgnk" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe> <p data-svelte-h="svelte-1qlvl3y">자동 음성 인식(Automatic Speech Recognition, ASR)은 음성 신호를 텍스트로 변환하여 음성 입력 시퀀스를 텍스트 출력에 매핑합니다.
	Siri와 Alexa와 같은 가상 어시스턴트는 ASR 모델을 사용하여 일상적으로 사용자를 돕고 있으며, 회의 중 라이브 캡션 및 메모 작성과 같은 유용한 사용자 친화적 응용 프로그램도 많이 있습니다.</p> <p data-svelte-h="svelte-7oxnvp">이 가이드에서 소개할 내용은 아래와 같습니다:</p> <ol data-svelte-h="svelte-1ok1jy5"><li><a href="https://huggingface.co/datasets/PolyAI/minds14" rel="nofollow">MInDS-14</a> 데이터 세트에서 <a href="https://huggingface.co/facebook/wav2vec2-base" rel="nofollow">Wav2Vec2</a>를 미세 조정하여 오디오를 텍스트로 변환합니다.</li> <li>미세 조정한 모델을 추론에 사용합니다.</li></ol> <blockquote class="tip"><p data-svelte-h="svelte-19nd5l3">이 작업과 호환되는 모든 아키텍처와 체크포인트를 보려면 <a href="https://huggingface.co/tasks/automatic-speech-recognition" rel="nofollow">작업 페이지</a>를 확인하는 것이 좋습니다.</p></blockquote> <p data-svelte-h="svelte-18iigii">시작하기 전에 필요한 모든 라이브러리가 설치되어 있는지 확인하세요:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-bash "><!-- HTML_TAG_START -->pip install transformers datasets evaluate jiwer<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-2mor9q">Hugging Face 계정에 로그인하면 모델을 업로드하고 커뮤니티에 공유할 수 있습니다. 토큰을 입력하여 로그인하세요.</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-py "><!-- HTML_TAG_START --><span class="hljs-meta">>>> </span><span class="hljs-keyword">from</span> huggingface_hub <span class="hljs-keyword">import</span> notebook_login

	<span class="hljs-meta">>>> </span>notebook_login()<!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="load-minds-14-dataset" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#load-minds-14-dataset"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>MInDS-14 데이터 세트 가져오기</span></h2> <p data-svelte-h="svelte-7q7xwh">먼저, 🤗 Datasets 라이브러리에서 <a href="https://huggingface.co/datasets/PolyAI/minds14" rel="nofollow">MInDS-14</a> 데이터 세트의 일부분을 가져오세요.
	이렇게 하면 전체 데이터 세트에 대한 훈련에 시간을 들이기 전에 모든 것이 작동하는지 실험하고 검증할 수 있습니다.</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-py "><!-- HTML_TAG_START --><span class="hljs-meta">>>> </span><span class="hljs-keyword">from</span> datasets <span class="hljs-keyword">import</span> load_dataset, Audio

	<span class="hljs-meta">>>> </span>minds = load_dataset(<span class="hljs-string">"PolyAI/minds14"</span>, name=<span class="hljs-string">"en-US"</span>, split=<span class="hljs-string">"train[:100]"</span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-3148pn"><code>~Dataset.train_test_split</code> 메소드를 사용하여 데이터 세트의 <code>train</code>을 훈련 세트와 테스트 세트로 나누세요:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-py "><!-- HTML_TAG_START --><span class="hljs-meta">>>> </span>minds = minds.train_test_split(test_size=<span class="hljs-number">0.2</span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-17o70r1">그리고 데이터 세트를 확인하세요:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-py "><!-- HTML_TAG_START --><span class="hljs-meta">>>> </span>minds
	DatasetDict({
	train: Dataset({
	features: [<span class="hljs-string">'path'</span>, <span class="hljs-string">'audio'</span>, <span class="hljs-string">'transcription'</span>, <span class="hljs-string">'english_transcription'</span>, <span class="hljs-string">'intent_class'</span>, <span class="hljs-string">'lang_id'</span>],
	num_rows: <span class="hljs-number">16</span>
	})
	test: Dataset({
	features: [<span class="hljs-string">'path'</span>, <span class="hljs-string">'audio'</span>, <span class="hljs-string">'transcription'</span>, <span class="hljs-string">'english_transcription'</span>, <span class="hljs-string">'intent_class'</span>, <span class="hljs-string">'lang_id'</span>],
	num_rows: <span class="hljs-number">4</span>
	})
	})<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1dp2qnq">데이터 세트에는 <code>lang_id</code>와 <code>english_transcription</code>과 같은 유용한 정보가 많이 포함되어 있지만, 이 가이드에서는 <code>audio</code>와 <code>transcription</code>에 초점을 맞출 것입니다. 다른 열은 <code>remove_columns</code> 메소드를 사용하여 제거하세요:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-py "><!-- HTML_TAG_START --><span class="hljs-meta">>>> </span>minds = minds.remove_columns([<span class="hljs-string">"english_transcription"</span>, <span class="hljs-string">"intent_class"</span>, <span class="hljs-string">"lang_id"</span>])<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1p9pg2d">예시를 다시 한번 확인해보세요:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-py "><!-- HTML_TAG_START --><span class="hljs-meta">>>> </span>minds[<span class="hljs-string">"train"</span>][<span class="hljs-number">0</span>]
	{<span class="hljs-string">'audio'</span>: {<span class="hljs-string">'array'</span>: array([-<span class="hljs-number">0.00024414</span>, <span class="hljs-number">0.</span> , <span class="hljs-number">0.</span> , ..., <span class="hljs-number">0.00024414</span>,
	<span class="hljs-number">0.00024414</span>, <span class="hljs-number">0.00024414</span>], dtype=float32),
	<span class="hljs-string">'path'</span>: <span class="hljs-string">'/root/.cache/huggingface/datasets/downloads/extracted/f14948e0e84be638dd7943ac36518a4cf3324e8b7aa331c5ab11541518e9368c/en-US~APP_ERROR/602ba9e2963e11ccd901cd4f.wav'</span>,
	<span class="hljs-string">'sampling_rate'</span>: <span class="hljs-number">8000</span>},
	<span class="hljs-string">'path'</span>: <span class="hljs-string">'/root/.cache/huggingface/datasets/downloads/extracted/f14948e0e84be638dd7943ac36518a4cf3324e8b7aa331c5ab11541518e9368c/en-US~APP_ERROR/602ba9e2963e11ccd901cd4f.wav'</span>,
	<span class="hljs-string">'transcription'</span>: <span class="hljs-string">"hi I'm trying to use the banking app on my phone and currently my checking and savings account balance is not refreshing"</span>}<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-6aiy91">두 개의 필드가 있습니다:</p> <ul data-svelte-h="svelte-1l53gx0"><li><code>audio</code>: 오디오 파일을 가져오고 리샘플링하기 위해 호출해야 하는 음성 신호의 1차원 <code>array(배열)</code></li> <li><code>transcription</code>: 목표 텍스트</li></ul> <h2 class="relative group"><a id="preprocess" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#preprocess"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>전처리</span></h2> <p data-svelte-h="svelte-1fm7wkp">다음으로 오디오 신호를 처리하기 위한 Wav2Vec2 프로세서를 가져옵니다:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-py "><!-- HTML_TAG_START --><span class="hljs-meta">>>> </span><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> AutoProcessor

	<span class="hljs-meta">>>> </span>processor = AutoProcessor.from_pretrained(<span class="hljs-string">"facebook/wav2vec2-base"</span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-e2ss04">MInDS-14 데이터 세트의 샘플링 레이트는 8000kHz이므로(<a href="https://huggingface.co/datasets/PolyAI/minds14" rel="nofollow">데이터 세트 카드</a>에서 확인), 사전 훈련된 Wav2Vec2 모델을 사용하려면 데이터 세트를 16000kHz로 리샘플링해야 합니다:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-py "><!-- HTML_TAG_START --><span class="hljs-meta">>>> </span>minds = minds.cast_column(<span class="hljs-string">"audio"</span>, Audio(sampling_rate=<span class="hljs-number">16_000</span>))
	<span class="hljs-meta">>>> </span>minds[<span class="hljs-string">"train"</span>][<span class="hljs-number">0</span>]
	{<span class="hljs-string">'audio'</span>: {<span class="hljs-string">'array'</span>: array([-<span class="hljs-number">2.38064706e-04</span>, -<span class="hljs-number">1.58618059e-04</span>, -<span class="hljs-number">5.43987835e-06</span>, ...,
	<span class="hljs-number">2.78103951e-04</span>, <span class="hljs-number">2.38446111e-04</span>, <span class="hljs-number">1.18740834e-04</span>], dtype=float32),
	<span class="hljs-string">'path'</span>: <span class="hljs-string">'/root/.cache/huggingface/datasets/downloads/extracted/f14948e0e84be638dd7943ac36518a4cf3324e8b7aa331c5ab11541518e9368c/en-US~APP_ERROR/602ba9e2963e11ccd901cd4f.wav'</span>,
	<span class="hljs-string">'sampling_rate'</span>: <span class="hljs-number">16000</span>},
	<span class="hljs-string">'path'</span>: <span class="hljs-string">'/root/.cache/huggingface/datasets/downloads/extracted/f14948e0e84be638dd7943ac36518a4cf3324e8b7aa331c5ab11541518e9368c/en-US~APP_ERROR/602ba9e2963e11ccd901cd4f.wav'</span>,
	<span class="hljs-string">'transcription'</span>: <span class="hljs-string">"hi I'm trying to use the banking app on my phone and currently my checking and savings account balance is not refreshing"</span>}<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-14tnwq5">위의 ‘transcription’에서 볼 수 있듯이 텍스트는 대문자와 소문자가 섞여 있습니다. Wav2Vec2 토크나이저는 대문자 문자에 대해서만 훈련되어 있으므로 텍스트가 토크나이저의 어휘와 일치하는지 확인해야 합니다:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-py "><!-- HTML_TAG_START --><span class="hljs-meta">>>> </span><span class="hljs-keyword">def</span> <span class="hljs-title function_">uppercase</span>(<span class="hljs-params">example</span>):
	<span class="hljs-meta">... </span> <span class="hljs-keyword">return</span> {<span class="hljs-string">"transcription"</span>: example[<span class="hljs-string">"transcription"</span>].upper()}


	<span class="hljs-meta">>>> </span>minds = minds.<span class="hljs-built_in">map</span>(uppercase)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-10u1mcw">이제 다음 작업을 수행할 전처리 함수를 만들어보겠습니다:</p> <ol data-svelte-h="svelte-bj7uax"><li><code>audio</code> 열을 호출하여 오디오 파일을 가져오고 리샘플링합니다.</li> <li>오디오 파일에서 <code>input_values</code>를 추출하고 프로세서로 <code>transcription</code> 열을 토큰화합니다.</li></ol> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-py "><!-- HTML_TAG_START --><span class="hljs-meta">>>> </span><span class="hljs-keyword">def</span> <span class="hljs-title function_">prepare_dataset</span>(<span class="hljs-params">batch</span>):
	<span class="hljs-meta">... </span> audio = batch[<span class="hljs-string">"audio"</span>]
	<span class="hljs-meta">... </span> batch = processor(audio[<span class="hljs-string">"array"</span>], sampling_rate=audio[<span class="hljs-string">"sampling_rate"</span>], text=batch[<span class="hljs-string">"transcription"</span>])
	<span class="hljs-meta">... </span> batch[<span class="hljs-string">"input_length"</span>] = <span class="hljs-built_in">len</span>(batch[<span class="hljs-string">"input_values"</span>][<span class="hljs-number">0</span>])
	<span class="hljs-meta">... </span> <span class="hljs-keyword">return</span> batch<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-vxgjmg">전체 데이터 세트에 전처리 함수를 적용하려면 🤗 Datasets <code>map</code> 함수를 사용하세요. <code>num_proc</code> 매개변수를 사용하여 프로세스 수를 늘리면 <code>map</code>의 속도를 높일 수 있습니다. <code>remove_columns</code> 메소드를 사용하여 필요하지 않은 열을 제거하세요:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-py "><!-- HTML_TAG_START --><span class="hljs-meta">>>> </span>encoded_minds = minds.<span class="hljs-built_in">map</span>(prepare_dataset, remove_columns=minds.column_names[<span class="hljs-string">"train"</span>], num_proc=<span class="hljs-number">4</span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-y3eg0c">🤗 Transformers에는 자동 음성 인식용 데이터 콜레이터가 없으므로 예제 배치를 생성하려면 <a href="/docs/transformers/main/ko/main_classes/data_collator#transformers.DataCollatorWithPadding">DataCollatorWithPadding</a>을 조정해야 합니다. 이렇게 하면 데이터 콜레이터는 텍스트와 레이블을 배치에서 가장 긴 요소의 길이에 동적으로 패딩하여 길이를 균일하게 합니다. <code>tokenizer</code> 함수에서 <code>padding=True</code>를 설정하여 텍스트를 패딩할 수 있지만, 동적 패딩이 더 효율적입니다.</p> <p data-svelte-h="svelte-pcjp4m">다른 데이터 콜레이터와 달리 이 특정 데이터 콜레이터는 <code>input_values</code>와 <code>labels</code>에 대해 다른 패딩 방법을 적용해야 합니다.</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-py "><!-- HTML_TAG_START --><span class="hljs-meta">>>> </span><span class="hljs-keyword">import</span> torch

	<span class="hljs-meta">>>> </span><span class="hljs-keyword">from</span> dataclasses <span class="hljs-keyword">import</span> dataclass, field
	<span class="hljs-meta">>>> </span><span class="hljs-keyword">from</span> typing <span class="hljs-keyword">import</span> <span class="hljs-type">Any</span>, <span class="hljs-type">Dict</span>, <span class="hljs-type">List</span>, <span class="hljs-type">Optional</span>, <span class="hljs-type">Union</span>


	<span class="hljs-meta">>>> </span>@dataclass
	<span class="hljs-meta">... </span><span class="hljs-keyword">class</span> <span class="hljs-title class_">DataCollatorCTCWithPadding</span>:
	<span class="hljs-meta">... </span> processor: AutoProcessor
	<span class="hljs-meta">... </span> padding: <span class="hljs-type">Union</span>[<span class="hljs-built_in">bool</span>, <span class="hljs-built_in">str</span>] = <span class="hljs-string">"longest"</span>

	<span class="hljs-meta">... </span> <span class="hljs-keyword">def</span> <span class="hljs-title function_">__call__</span>(<span class="hljs-params">self, features: <span class="hljs-built_in">list</span>[<span class="hljs-built_in">dict</span>[<span class="hljs-built_in">str</span>, <span class="hljs-type">Union</span>[<span class="hljs-built_in">list</span>[<span class="hljs-built_in">int</span>], torch.Tensor]]]</span>) -> <span class="hljs-built_in">dict</span>[<span class="hljs-built_in">str</span>, torch.Tensor]:
	<span class="hljs-meta">... </span> <span class="hljs-comment"># 입력과 레이블을 분할합니다</span>
	<span class="hljs-meta">... </span> <span class="hljs-comment"># 길이가 다르고, 각각 다른 패딩 방법을 사용해야 하기 때문입니다</span>
	<span class="hljs-meta">... </span> input_features = [{<span class="hljs-string">"input_values"</span>: feature[<span class="hljs-string">"input_values"</span>][<span class="hljs-number">0</span>]} <span class="hljs-keyword">for</span> feature <span class="hljs-keyword">in</span> features]
	<span class="hljs-meta">... </span> label_features = [{<span class="hljs-string">"input_ids"</span>: feature[<span class="hljs-string">"labels"</span>]} <span class="hljs-keyword">for</span> feature <span class="hljs-keyword">in</span> features]

	<span class="hljs-meta">... </span> batch = self.processor.pad(input_features, padding=self.padding, return_tensors=<span class="hljs-string">"pt"</span>)

	<span class="hljs-meta">... </span> labels_batch = self.processor.pad(labels=label_features, padding=self.padding, return_tensors=<span class="hljs-string">"pt"</span>)

	<span class="hljs-meta">... </span> <span class="hljs-comment"># 패딩에 대해 손실을 적용하지 않도록 -100으로 대체합니다</span>
	<span class="hljs-meta">... </span> labels = labels_batch[<span class="hljs-string">"input_ids"</span>].masked_fill(labels_batch.attention_mask.ne(<span class="hljs-number">1</span>), -<span class="hljs-number">100</span>)

	<span class="hljs-meta">... </span> batch[<span class="hljs-string">"labels"</span>] = labels

	<span class="hljs-meta">... </span> <span class="hljs-keyword">return</span> batch<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-o882aa">이제 <code>DataCollatorForCTCWithPadding</code>을 인스턴스화합니다:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-py "><!-- HTML_TAG_START --><span class="hljs-meta">>>> </span>data_collator = DataCollatorCTCWithPadding(processor=processor, padding=<span class="hljs-string">"longest"</span>)<!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="evaluate" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#evaluate"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>평가하기</span></h2> <p data-svelte-h="svelte-1kx52w5">훈련 중에 평가 지표를 포함하면 모델의 성능을 평가하는 데 도움이 되는 경우가 많습니다. 🤗 <a href="https://huggingface.co/docs/evaluate/index" rel="nofollow">Evaluate</a> 라이브러리를 사용하면 평가 방법을 빠르게 불러올 수 있습니다.
	이 작업에서는 <a href="https://huggingface.co/spaces/evaluate-metric/wer" rel="nofollow">단어 오류율(Word Error Rate, WER)</a> 평가 지표를 가져옵니다.
	(평가 지표를 불러오고 계산하는 방법은 🤗 Evaluate <a href="https://huggingface.co/docs/evaluate/a_quick_tour" rel="nofollow">둘러보기</a>를 참조하세요):</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-py "><!-- HTML_TAG_START --><span class="hljs-meta">>>> </span><span class="hljs-keyword">import</span> evaluate

	<span class="hljs-meta">>>> </span>wer = evaluate.load(<span class="hljs-string">"wer"</span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-i1fso8">그런 다음 예측값과 레이블을 <code>compute</code>에 전달하여 WER을 계산하는 함수를 만듭니다:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-py "><!-- HTML_TAG_START --><span class="hljs-meta">>>> </span><span class="hljs-keyword">import</span> numpy <span class="hljs-keyword">as</span> np


	<span class="hljs-meta">>>> </span><span class="hljs-keyword">def</span> <span class="hljs-title function_">compute_metrics</span>(<span class="hljs-params">pred</span>):
	<span class="hljs-meta">... </span> pred_logits = pred.predictions
	<span class="hljs-meta">... </span> pred_ids = np.argmax(pred_logits, axis=-<span class="hljs-number">1</span>)

	<span class="hljs-meta">... </span> pred.label_ids[pred.label_ids == -<span class="hljs-number">100</span>] = processor.tokenizer.pad_token_id

	<span class="hljs-meta">... </span> pred_str = processor.batch_decode(pred_ids)
	<span class="hljs-meta">... </span> label_str = processor.batch_decode(pred.label_ids, group_tokens=<span class="hljs-literal">False</span>)

	<span class="hljs-meta">... </span> wer = wer.compute(predictions=pred_str, references=label_str)

	<span class="hljs-meta">... </span> <span class="hljs-keyword">return</span> {<span class="hljs-string">"wer"</span>: wer}<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-15lq6yk">이제 <code>compute_metrics</code> 함수를 사용할 준비가 되었으며, 훈련을 설정할 때 이 함수로 되돌아올 것입니다.</p> <h2 class="relative group"><a id="train" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#train"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>훈련하기</span></h2> <blockquote class="tip"><p data-svelte-h="svelte-1qq4it9"><a href="/docs/transformers/main/ko/main_classes/trainer#transformers.Trainer">Trainer</a>로 모델을 미세 조정하는 것이 익숙하지 않다면, <a href="../training#train-with-pytorch-trainer">여기</a>에서 기본 튜토리얼을 확인해보세요!</p></blockquote> <p data-svelte-h="svelte-1necmh2">이제 모델 훈련을 시작할 준비가 되었습니다! <a href="/docs/transformers/main/ko/model_doc/auto#transformers.AutoModelForCTC">AutoModelForCTC</a>로 Wav2Vec2를 가져오세요. <code>ctc_loss_reduction</code> 매개변수로 CTC 손실에 적용할 축소(reduction) 방법을 지정하세요. 기본값인 합계 대신 평균을 사용하는 것이 더 좋은 경우가 많습니다:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-py "><!-- HTML_TAG_START --><span class="hljs-meta">>>> </span><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> AutoModelForCTC, TrainingArguments, Trainer

	<span class="hljs-meta">>>> </span>model = AutoModelForCTC.from_pretrained(
	<span class="hljs-meta">... </span> <span class="hljs-string">"facebook/wav2vec2-base"</span>,
	<span class="hljs-meta">... </span> ctc_loss_reduction=<span class="hljs-string">"mean"</span>,
	<span class="hljs-meta">... </span> pad_token_id=processor.tokenizer.pad_token_id,
	<span class="hljs-meta">... </span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1vwg7jz">이제 세 단계만 남았습니다:</p> <ol data-svelte-h="svelte-xhsnjh"><li><a href="/docs/transformers/main/ko/main_classes/trainer#transformers.TrainingArguments">TrainingArguments</a>에서 훈련 하이퍼파라미터를 정의하세요. <code>output_dir</code>은 모델을 저장할 경로를 지정하는 유일한 필수 매개변수입니다. <code>push_to_hub=True</code>를 설정하여 모델을 Hub에 업로드 할 수 있습니다(모델을 업로드하려면 Hugging Face에 로그인해야 합니다). <a href="/docs/transformers/main/ko/main_classes/trainer#transformers.Trainer">Trainer</a>는 각 에폭마다 WER을 평가하고 훈련 체크포인트를 저장합니다.</li> <li>모델, 데이터 세트, 토크나이저, 데이터 콜레이터, <code>compute_metrics</code> 함수와 함께 <a href="/docs/transformers/main/ko/main_classes/trainer#transformers.Trainer">Trainer</a>에 훈련 인수를 전달하세요.</li> <li><a href="/docs/transformers/main/ko/main_classes/trainer#transformers.Trainer.train">train()</a>을 호출하여 모델을 미세 조정하세요.</li></ol> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-py "><!-- HTML_TAG_START --><span class="hljs-meta">>>> </span>training_args = TrainingArguments(
	<span class="hljs-meta">... </span> output_dir=<span class="hljs-string">"my_awesome_asr_mind_model"</span>,
	<span class="hljs-meta">... </span> per_device_train_batch_size=<span class="hljs-number">8</span>,
	<span class="hljs-meta">... </span> gradient_accumulation_steps=<span class="hljs-number">2</span>,
	<span class="hljs-meta">... </span> learning_rate=<span class="hljs-number">1e-5</span>,
	<span class="hljs-meta">... </span> warmup_steps=<span class="hljs-number">500</span>,
	<span class="hljs-meta">... </span> max_steps=<span class="hljs-number">2000</span>,
	<span class="hljs-meta">... </span> gradient_checkpointing=<span class="hljs-literal">True</span>,
	<span class="hljs-meta">... </span> fp16=<span class="hljs-literal">True</span>,
	<span class="hljs-meta">... </span> train_sampling_strategy=<span class="hljs-string">"group_by_length"</span>,
	<span class="hljs-meta">... </span> eval_strategy=<span class="hljs-string">"steps"</span>,
	<span class="hljs-meta">... </span> per_device_eval_batch_size=<span class="hljs-number">8</span>,
	<span class="hljs-meta">... </span> save_steps=<span class="hljs-number">1000</span>,
	<span class="hljs-meta">... </span> eval_steps=<span class="hljs-number">1000</span>,
	<span class="hljs-meta">... </span> logging_steps=<span class="hljs-number">25</span>,
	<span class="hljs-meta">... </span> load_best_model_at_end=<span class="hljs-literal">True</span>,
	<span class="hljs-meta">... </span> metric_for_best_model=<span class="hljs-string">"wer"</span>,
	<span class="hljs-meta">... </span> greater_is_better=<span class="hljs-literal">False</span>,
	<span class="hljs-meta">... </span> push_to_hub=<span class="hljs-literal">True</span>,
	<span class="hljs-meta">... </span>)

	<span class="hljs-meta">>>> </span>trainer = Trainer(
	<span class="hljs-meta">... </span> model=model,
	<span class="hljs-meta">... </span> args=training_args,
	<span class="hljs-meta">... </span> train_dataset=encoded_minds[<span class="hljs-string">"train"</span>],
	<span class="hljs-meta">... </span> eval_dataset=encoded_minds[<span class="hljs-string">"test"</span>],
	<span class="hljs-meta">... </span> processing_class=processor.feature_extractor,
	<span class="hljs-meta">... </span> data_collator=data_collator,
	<span class="hljs-meta">... </span> compute_metrics=compute_metrics,
	<span class="hljs-meta">... </span>)

	<span class="hljs-meta">>>> </span>trainer.train()<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1cjvbx4">훈련이 완료되면 모두가 모델을 사용할 수 있도록 <a href="/docs/transformers/main/ko/main_classes/trainer#transformers.Trainer.push_to_hub">push_to_hub()</a> 메소드를 사용하여 모델을 Hub에 공유하세요:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-py "><!-- HTML_TAG_START --><span class="hljs-meta">>>> </span>trainer.push_to_hub()<!-- HTML_TAG_END --></pre></div> <blockquote class="tip"><p data-svelte-h="svelte-1aqt9cb">자동 음성 인식을 위해 모델을 미세 조정하는 더 자세한 예제는 영어 자동 음성 인식을 위한 <a href="https://huggingface.co/blog/fine-tune-wav2vec2-english" rel="nofollow">블로그 포스트</a>와 다국어 자동 음성 인식을 위한 <a href="https://huggingface.co/blog/fine-tune-xlsr-wav2vec2" rel="nofollow">포스트</a>를 참조하세요.</p></blockquote> <h2 class="relative group"><a id="inference" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#inference"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>추론하기</span></h2> <p data-svelte-h="svelte-1annmqf">좋아요, 이제 모델을 미세 조정했으니 추론에 사용할 수 있습니다!</p> <p data-svelte-h="svelte-18ieb63">추론에 사용할 오디오 파일을 가져오세요. 필요한 경우 오디오 파일의 샘플링 비율을 모델의 샘플링 레이트에 맞게 리샘플링하는 것을 잊지 마세요!</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-py "><!-- HTML_TAG_START --><span class="hljs-meta">>>> </span><span class="hljs-keyword">from</span> datasets <span class="hljs-keyword">import</span> load_dataset, Audio

	<span class="hljs-meta">>>> </span>dataset = load_dataset(<span class="hljs-string">"PolyAI/minds14"</span>, <span class="hljs-string">"en-US"</span>, split=<span class="hljs-string">"train"</span>)
	<span class="hljs-meta">>>> </span>dataset = dataset.cast_column(<span class="hljs-string">"audio"</span>, Audio(sampling_rate=<span class="hljs-number">16000</span>))
	<span class="hljs-meta">>>> </span>sampling_rate = dataset.features[<span class="hljs-string">"audio"</span>].sampling_rate
	<span class="hljs-meta">>>> </span>audio_file = dataset[<span class="hljs-number">0</span>][<span class="hljs-string">"audio"</span>][<span class="hljs-string">"path"</span>]<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-12cj03s">추론을 위해 미세 조정된 모델을 시험해보는 가장 간단한 방법은 <code>pipeline()</code>을 사용하는 것입니다. 모델을 사용하여 자동 음성 인식을 위한 <code>pipeline</code>을 인스턴스화하고 오디오 파일을 전달하세요:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-py "><!-- HTML_TAG_START --><span class="hljs-meta">>>> </span><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> pipeline

	<span class="hljs-meta">>>> </span>transcriber = pipeline(<span class="hljs-string">"automatic-speech-recognition"</span>, model=<span class="hljs-string">"stevhliu/my_awesome_asr_minds_model"</span>)
	<span class="hljs-meta">>>> </span>transcriber(audio_file)
	{<span class="hljs-string">'text'</span>: <span class="hljs-string">'I WOUD LIKE O SET UP JOINT ACOUNT WTH Y PARTNER'</span>}<!-- HTML_TAG_END --></pre></div> <blockquote class="tip"><p data-svelte-h="svelte-1bgb7bv">텍스트로 변환된 결과가 꽤 괜찮지만 더 좋을 수도 있습니다! 더 나은 결과를 얻으려면 더 많은 예제로 모델을 미세 조정하세요!</p></blockquote> <p data-svelte-h="svelte-1olk0w9"><code>pipeline</code>의 결과를 수동으로 재현할 수도 있습니다:</p> <p data-svelte-h="svelte-1l4sx2f">오디오 파일과 텍스트를 전처리하고 PyTorch 텐서로 <code>input</code>을 반환할 프로세서를 가져오세요:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-py "><!-- HTML_TAG_START --><span class="hljs-meta">>>> </span><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> AutoProcessor

	<span class="hljs-meta">>>> </span>processor = AutoProcessor.from_pretrained(<span class="hljs-string">"stevhliu/my_awesome_asr_mind_model"</span>)
	<span class="hljs-meta">>>> </span>inputs = processor(dataset[<span class="hljs-number">0</span>][<span class="hljs-string">"audio"</span>][<span class="hljs-string">"array"</span>], sampling_rate=sampling_rate, return_tensors=<span class="hljs-string">"pt"</span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-2w5h5p">입력을 모델에 전달하고 로짓을 반환하세요:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-py "><!-- HTML_TAG_START --><span class="hljs-meta">>>> </span><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> AutoModelForCTC

	<span class="hljs-meta">>>> </span>model = AutoModelForCTC.from_pretrained(<span class="hljs-string">"stevhliu/my_awesome_asr_mind_model"</span>)
	<span class="hljs-meta">>>> </span><span class="hljs-keyword">with</span> torch.no_grad():
	<span class="hljs-meta">... </span> logits = model(**inputs).logits<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-v9dw3r">가장 높은 확률의 <code>input_ids</code>를 예측하고, 프로세서를 사용하여 예측된 <code>input_ids</code>를 다시 텍스트로 디코딩하세요:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-py "><!-- HTML_TAG_START --><span class="hljs-meta">>>> </span><span class="hljs-keyword">import</span> torch

	<span class="hljs-meta">>>> </span>predicted_ids = torch.argmax(logits, dim=-<span class="hljs-number">1</span>)
	<span class="hljs-meta">>>> </span>transcription = processor.batch_decode(predicted_ids)
	<span class="hljs-meta">>>> </span>transcription
	[<span class="hljs-string">'I WOUL LIKE O SET UP JOINT ACOUNT WTH Y PARTNER'</span>]<!-- HTML_TAG_END --></pre></div> <a class="!text-gray-400 !no-underline text-sm flex items-center not-prose mt-4" href="https://github.com/huggingface/transformers/blob/main/docs/source/ko/tasks/asr.md" target="_blank"><svg class="mr-1" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M31,16l-7,7l-1.41-1.41L28.17,16l-5.58-5.59L24,9l7,7z"></path><path d="M1,16l7-7l1.41,1.41L3.83,16l5.58,5.59L8,23l-7-7z"></path><path d="M12.419,25.484L17.639,6.552l1.932,0.518L14.351,26.002z"></path></svg> <span data-svelte-h="svelte-zjs2n5"><span class="underline">Update</span> on GitHub</span></a> <p></p>

	<script>
	{
	__sveltekit_16kp15g = {
	assets: "/docs/transformers/main/ko",
	base: "/docs/transformers/main/ko",
	env: {}
	};

	const element = document.currentScript.parentElement;

	const data = [null,null];

	Promise.all([
	import("/docs/transformers/main/ko/_app/immutable/entry/start.acfca1c5.js"),
	import("/docs/transformers/main/ko/_app/immutable/entry/app.f0c7a87a.js")
	]).then(([kit, app]) => {
	kit.start(app, element, {
	node_ids: [0, 155],
	data,
	form: null,
	error: null
	});
	});
	}
	</script>

Xet Storage Details

Size:: 73.2 kB
Xet hash:: 47ccdaaf9a150111f7b3aebcc97de99420a4f61436b6b4d37018c064cff1bc4b

Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.